Ph.D. in progress · Instituto Superior Técnico & Carnegie Mellon University
Lisbon, Portugal
My research spans neural information retrieval, learned sparse representations, passage re-ranking, and natural language processing. Each publication below includes a short summary and links to the paper; code and data are included where available.
See also my Google Scholar profile.
Luís Borges, Rohan Jha, Jamie Callan, and Bruno Martins
ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR)
Combining retrieval and LLM re-ranking to find remembered items from uncertain descriptions across multiple domains.
Luís Borges, Jamie Callan, and Bruno Martins
Text Retrieval Conference (TREC), participant paper
A two-stage movie retrieval system combining dense passage retrieval with zero-shot LLM re-ranking for uncertain, verbose queries.
Luís Borges, Bruno Martins, and Jamie Callan
ACM SIGIR International Conference on Theory of Information Retrieval (ICTIR)
Learning sparse representations over a latent vocabulary to improve neural search while retaining inverted index retrieval.
Luís Borges, Bruno Martins, and Jamie Callan
European Conference on Information Retrieval (ECIR)
Evaluating model ensembles, rank fusion, and learning-to-rank for more effective neural passage re-ranking on MS MARCO.
Luís Borges, Bruno Martins, and Pável Calado
ACM Journal of Data and Information Quality, 11(3)
Combining neural text representations, similarity features, and transfer learning to classify the stance of news articles toward headlines.